In [28]:
# calculate the correlations
correlations = merged_df_final.corr()
# list of all features except the target
all_features = merged_df_final.columns.tolist()
target_variable = 'critical_temp_x'
all_features.remove(target_variable)
#Define the number of features per group
features_per_group = 4
total_features = len(all_features)
total_groups = (total_features + features_per_group - 1) // features_per_group
# Iterate through groups
for group_num in range(total_groups):
start_idx = group_num * features_per_group
end_idx = min((group_num + 1) * features_per_group, total_features)
selected_features = all_features[start_idx:end_idx]
#Create a new figure for each group
plt.figure(figsize=(15,10))
for idx, feature in enumerate(selected_features):
plt.subplot(2, 4, idx + 1)
sns.scatterplot(data=merged_df_final, x=feature, y=target_variable)
# Calcualte correlation coefficient
corr_coeff = correlations.loc[feature, target_variable]
#Annoate twith correlation coefficients
plt.text(0.5, 0.9, f'Corr: {corr_coeff: .2f}', horizontalalignment='center',
verticalalignment='center',transform=plt.gca().transAxes, fontsize=10)
plt.title(f"Scatter Plot: {target_variable} vs {feature}", fontsize = 9) #adjust title size here
plt.tight_layout()
plt.subplots_adjust(top=0.9) #Adjust top spacing for the overall title
plt.suptitle(f"Correlation Plots - Group {group_num+1}", fontsize = 16) # Overalltitle
plt.show()
